Agent interfaces are becoming increasingly sophisticated in their ability to allow access to numerous types of application data and/or application systems across multiple forms of communication. For example, a typical customer service application may include an agent interface to allow a customer service agent to navigate among a variety of types of data related to a customer and to products. Such product data may include a knowledge base or other database of product information, while customer data may include contact information, service request information, order information, activity information, and so on. A customer service agent interacting with a set of customers may need to navigate quickly all of these types of information during, for example, the course of a parallel set of chat conversations involving several different customers in separate chat (or other pseudo-real-time communication) sessions.
In order to provide prompt and personalized service when processing the needs of several customers in parallel, it is desirable that customer service agents appear to “know” the customer immediately when starting an interaction and throughout its duration. Since customer service centers receive a large volume and variety of customer interactions during a typical day, each agent needs to have quick access to important customer information, such as contact name, account number, phone number, and so on, for the customer with whom the agent is dealing. Additionally, customer service agents need to find, in the most efficient manner possible, the information needed by the customer. When an agent wastes time by fumbling with an agent interface for a search engine, suggesting to a customer an inapplicable resource, or offering otherwise inapplicable advice, costs are increased. These costs come in both the form of agent time and the form of customer dissatisfaction.